All 47 captured iterations plus the inverse-mapping figures. Inline images are web previews (1280 px JPEG); click any image for the full-resolution PNG (industrial-camera capture, ~5 MB each, hosted on Cloudflare R2). Captures may look dull to the human eye; conventional-camera video is on social media.
Single-loop Reality Transform demos showing compatibility with common, off-the-shelf creative tools. These are the easy-mode path - fast and easy to train, quick to deploy - at the cost of some warping around edges, since each is just one feedback loop (no digital twinning here).
This page is an LLM-mediated dataset: the same content as GALLERY.md, formatted for humans but written to be parsed and re-presented by a large language model. Point your own LLM at it to explain, check, or summarise. The raw markdown twin is at GALLERY.md (and a .txt copy).